This invention relates to an information retrieving system and more particularly to a signal retrieving system used for reading an optical recording medium such as an optical disc or an optical tape and suitable for retrieving data stably and accurately irrespective of a signal modulation system and variations of a reflection factor.
A method which utilizes a servo byte area in order to carry out data modulation/demodulation, focussing servo and tracking servo is discussed in SPIE Proceedings, Vol. 695, Optical Mass Data Storage 2 (1986). This article does not particularly mention a data modulation scheme but a data modulation scheme whose mean value does not change easily in accordance with the density of data patterns and which has high direct current free characteristics is recommended at present.
Generally, a data modulation system is evaluated by self-clocking characteristics, a density ratio and data detection window width in addition to the direct current free characteristics. An MFM (Modified FM) scheme used for a magnetic disc, for example, generates "1" on the rear side of a bit cell in response to the data bit "1" and generates "1" on the front side of the bit cell with respect to the second "0" et seq if two or more "0"s continue. Therefore, inversion of "0" and "1" occurs at a suitable frequency in a recording pulse even when "0" or "1" continues in the data bit. Therefore, the direct current free characteristics are high and moreover, a demodulation clock can be generated from the recorded data stream itself. In other words, self-clocking can be accomplished. On the other hand, it is necessary to examine the existence of the recording bit for the 1/2 detection window width of the data bit width in order to demodulate correctly.
NRZ (Non-Return to Zero) modulation is a scheme which records "1" and "0" of the data bit as the code word, and does not have the direct current free characteristics and the self-clocking ability. However, since the detection window width is the same as the data bit width, this scheme has resistance to jitter fluctuation.
According to a sampled servo format or a format of the same kind, a modulation system can be determined irrespective of the existence of the self-clocking ability of the data itself. As to the direct current free characteristics, however, demodulation in an AC-coupled amplification system is difficult. In the cases where a considerable number of "0"s continue and where a considerable number of "1"s continue, for example, the signal levels whose DC component is cut becomes the same in the AC-coupled amplification system so that if digitization is made at a predetermined threshold value, it becomes impossible to discriminate between "0" and "1". If a system having the direct current free characteristics is used as the modulation system, variations of the reproduced signal level occur due to variations of a reflection factor of the recording medium and retardation.
As described above, in order to obtain correct data from the optical reproduction signal, it is necessary to improve the direct current free characteristics and to suppress the variation of the reproduction signal itself. It is also necessary to suppress the waveform interference arising from the interference of the waveform of the reproduction signal with an adjacent pit and asymmetry of the reproduction signal arising from non-uniformity of pit shapes. In the case of a circular recording pit, for example, asymmetry arising from the pit shape non-uniformity is less and the influence of the waveforms that occur when the adjacent pits come close to one another is rather great. On the other hand, in the case of edge recording for forming an elliptic recording pit by making the leading and trailing edges of the pit correspond to code word "1", the influence of asymmetry arising from the pit shape non-uniformity is great. Since the recording pit is formed by thermal recording, the width of the trailing edge of the recording pit becomes greater than the width of the leading edge. Therefore, the inclination of the reproduction waveform differs between the leading edge and the trailing edge. Therefore, if interpretation of a signal is made at a predetermined threshold value and the signal is converted to a digital signal, the edge position fluctuates.